Munich Personal RePEc Archive

Multivariate GARCH Approaches: case of major sectorial Tunisian stock markets

NEIFAR, MALIKA (2020): Multivariate GARCH Approaches: case of major sectorial Tunisian stock markets.

[thumbnail of MPRA_paper_99658.pdf]

Download (2MB) | Preview


The objectif in this paper is to proposes multivariate GARCH volatility models to assess the dynamic interdependence among volatility of returns for 5 tunisian sectorial stock index series (namely : Bank, FINancial service, AUTOmobile, INDustry, and Materials (MATB)) and TUNindex series. The Monthly returns of stock indices have been considered from 2010M02 to 2019M07. Two systems are considered. The first System, with Constant Conditional (C) mean, allows for market interaction. Results from DVECH model reveals that some sectorial stock markets are interdependent, the presence of a significance and positive effect of cross shock of Finance and Bank stock returns on Tunindex return, and volatility is predictable. C Correlation, ρij, have decreasing evolution for full period or for recent years for almost all i and j except CC between Tunindex return and R_FIN (and R_BANK) and CC between R_FIN and R_IND (and R_MATB). The tests for volatility spillovers effects suggests significant volatility spillovers from MATB and AUTO sectors to IND sector and from AUTO sector to MATB sector. The second system, with macroeconomic factor instability effects as Conditional mean, examine the CCC and DCC between different sectors. The main result supports the hypotheses of DCC. The DCC provides evidence of cross border relationship between sectors and macro economic instability factors have significant effect on the mean of returns evolutions (at 5% or 10% level). Volatility of exchange rate has significant positive effect on R, R_FIN, and R_MATB, while volatility of inflation has significant negative effect on R_Fin and volatility of oil price has significant negative effect on R_AUTO.

Atom RSS 1.0 RSS 2.0

Contact us: mpra@ub.uni-muenchen.de

This repository has been built using EPrints software.

MPRA is a RePEc service hosted by Logo of the University Library LMU Munich.